Diagnostic Method for Signal Discontinuities in Engine Controllers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diagnostic methods for internal combustion engines face challenges in accurately detecting signal jumps due to component-related fluctuations, leading to inaccurate error detection and failure to recognize leaks or malfunctions, especially in systems with large component scattering and high signal variance.
Innovation Solution
A diagnostic method that involves collecting an input signal, checking a jump criterion, and releasing a diagnosis by setting a reference value and determining a relative diagnostic size, which is compared to a diagnostic threshold to detect faults, thereby increasing sensitivity and accurately identifying errors such as leaks or malfunctions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If diagnostic thresholds are set to detect component variations, then sensitivity to faults increases, but false fault detections increase due to normal component tolerances
Solution Approach 1:
The system performs preliminary learning of component-specific characteristics during a fault-free operation period before actual diagnostic evaluation begins. This preliminary action establishes a baseline model of normal component behavior, allowing the diagnostic system to distinguish between normal tolerances and actual faults more accurately.
Solution Approach 2:
The diagnostic threshold is not fixed but dynamically adapted based on learned component characteristics. The system continuously updates the threshold based on the specific component's behavior patterns, making the threshold sensitive to actual faults while tolerant of normal variations for that specific component.
2Reliability
If component-specific modeling is implemented to account for tolerances, then false detections decrease, but system complexity and modeling requirements increase
Solution Approach 1:
The system performs self-learning of component characteristics automatically during normal operation without requiring external calibration or manual modeling. The control unit autonomously collects data, identifies component-specific patterns, and adapts diagnostic thresholds, eliminating the need for complex manual modeling procedures.
Solution Approach 2:
Instead of creating complex structural models for each component, the system changes the parameters used for diagnosis by learning statistical characteristics (mean, standard deviation) of component behavior. This parameter-based approach simplifies the modeling requirement while maintaining component-specific accuracy.
3Device complexity
If global models are used for fault detection, then system simplicity is maintained, but detection accuracy decreases due to inability to account for component variations
Solution Approach 1:
The system starts with a simple global diagnostic model but dynamically adapts it by incorporating learned component-specific characteristics. The threshold evolves from a fixed global value to a dynamic, component-adjusted value, maintaining simplicity in the overall approach while improving accuracy through adaptive parameter changes.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A diagnostic method for detecting discontinuities in a continuous measurement variable, comprising: capturing an input signal (I), testing a discontinuity criterion (BD) on the basis of the input signal (I), when the discontinuity criterion (BD) is met, enabling a diagnosis, comprising: capturing the input signal (I) within a transit time (T) starting from a point in time (t0), defining the input signal (l0) at the point in time t0 of the transit time (T) as a reference value (R) starting from an initial reference position (R0), determining the profile of a relative diagnostic variable (IR) by shifting the profile of the input signal (I) by the reference difference (RD) between the initial reference position (R0) and the reference value (R), wherein the profile of the relative diagnostic variable (IR) is increased by the reference difference (RD) if the reference value (l0; R) is lower than the initial reference value (R0), and is decreased by the reference difference (RD) if the reference value (l0; R) is higher than the initial reference value (R0); comparing the profile of the relative diagnostic variable (IR) with a diagnostic threshold (D), and testing an error criterion (FK), according to which the relative diagnostic variable (IR) exceeds the diagnostic threshold (D). The invention further relates to a controller (9) for carrying out the diagnostic method and to a vehicle (100) having such a controller (9).